This proposal takes the initial steps in the development of a general model for predictive toxicology. The bioactivation of xenobiotics to reactive chemical species, by the CYPs (cytochrome P-450s), is recognized as an important pathway in the generation of various toxicities and in chemical carcinogenesis. In fact, it has been proposed that the majority of oxidative bioactivation reactions can be attributed to CYP. Experimental and theoretical methods will be used to describe quantitatively the factors involved in binding and catalysis of three distinct classes of compounds that can be bioactivated to reactive species. Knowledge of these factors will have significant impact on our ability to predict toxicity and will enhance our ability to understand the factors that modulate chemical carcinogenesis. This model will also clay an important role in reducing our dependence on animal models to predict human toxicity and help in the elucidation of the mechanisms of toxicity. In this proposal theoretical calculations of the energetics of various CYP catalyzed bioactivation reactions will be correlated with experimental data. Three classes of compounds will be studied; a) a series of nitriles b) a series of hydrochlorofluorocarbons, and c) a series of para substituted toluenes. Methods will be outlined for the determination of the relative binding constants and the rates of reaction of these substrates with both quantum chemical and molecular dynamics techniques. Preliminary results show that accurate predictions of the rates of reaction and binding energies can be made with the theoretical models described in this proposal. If this predictive method had been established prior to the initial development of the HCFC replacements for CFCs many research animals would have been spared and a large sum of money would have been saved on development costs. An excellent correlation of predicted rates of metabolism with in vivo toxicity has also been confirmed for the metabolism of nitrile containing compounds. New results indicate that the in vivo human metabolic rates of the inhalation anesthetics can be predicted using our AM1 model. We believe this to be the first prediction of the relative rate of metabolism in humans. Furthermore, we have been able to determine the stereochemistry and specific amino acid residues that are responsible for activating the procarcinogen benzo[a]pyrene to the most carcinogenic (7R,8R]-diol using P45Ocam and molecular dynamics.